Keyword Research

SEO: 8 Reasons to Stop the Addiction to Keyword Demand

Search marketing is all about experimentation — testing and learning. Slavish obsession with the average number of searches per month as a sole metric for optimization or a forecasting revenue will only hurt a natural search program.

It’s time to end our addiction to search demand.

The number of searches per month displayed in a keyword tool is a siren’s song. You see a huge number of searches and immediately think of how many could come to your site after you optimize a page.

Of the 2,617,550 searches that Google approximates in the U.S. last December for “smartwatch” and related keyword themes, it’s tempting to think, “Surely I can capture 1 percent and drive 260,000 visits this December! Or a tenth of a percent and drive 26,000 visitors.”

That’s a good thought. But what will you do with those visitors, however many you capture? Where do they land and where do they convert? Do they bounce out and spend their money someplace else? And do the numbers even mean what you think?

It’s easy to see search demand as potential revenue. There’s even a handy formula to get there.

Revenue forecasting formula

Natural search formula for computing revenue based on search demand.

If you compute that by replacing the items in the formula with sample numbers, it comes to something like this.

Formula for revenue forecasting with data.

Sample completed formula for computing revenue based on search demand.

However, calculating revenue from search demand is only a theory. There are problems when it’s applied in the real world.

8 Reasons to Stop the Addiction to Keyword Demand

  • One size does not fit all. Averages don’t apply evenly across keyword themes in natural search. If you drive .1 percent of your head keyword (301,000 for “smart watch”), and 10 percent of a longer tail keyword like “best smartwatch battery life” at 480 searches on average per month, the amount of value that those two drive summed together equals $1,711 in total revenue. If you apply the blanket formula above, you get $14,787. That’s a difference of almost a factor of 10.
  • You only see the head keyword themes, and only for the data that Google decides to show you. If you do enough keyword research, you’ll start to doubt the data. Researching one phrase may show 10 related keywords in a theme worth about 45,000 visits. But typing a second phrase — such as a plural version of the same keyword phrase — into the keyword input field could actually reduce the number of related keywords shown, even though the keywords are nearly identical.
  • All keyword theme numbers in Google Keyword Planner are estimates.
  • You only see the top traffic-driving keywords. It’s impossible to research the long tail. Those millions of keywords that people search for only once, or even five times per month, won’t show up in Google Keyword Planner. They can’t. There’s too much data represented in the trillions of searches that Google handles a year. So when you’re researching keywords, you’re only pulling from the top of the keyword data pile. How deep does it go? Only Google knows.
  • You’re pulling data from two different systems to get traffic data by keyword and revenue data for natural search because the data for the entire formula isn’t available in a single tool. And the data represented in two different systems never matches.
  • The average order value and conversion rate in your web analytics package take into account every scrap of natural search traffic, but the natural search click-through data in Google Search Console is for a small subset — just the top couple of thousand keywords. It’s highly unlikely that the comprehensive web analytics data exactly tracks with the much smaller Search Console numbers. That’s going to throw off your calculations.
  • The keyword data in Google Keyword Planner only represents searches conducted in Google. That means that the amount of demand — adding in other search engines — may be 10 to 15 percent higher than what you’re estimating. But applying an increase of a set amount is also not accurate because it will vary based on each keyword. How much is unknowable.
  • Past performance is no guarantee of future gains. Keyword demand fluctuates with seasonality and trends. Click rates vary based on rankings and Google’s frequent algorithm updates. Conversion rates change based on sudden modifications to the content or platform for the site. Average order value changes with the pricing of products and the mercurial passions of your customer base. With so many changeable factors, it only takes one slip in one area to throw the whole equation into motion and impact the revenue forecast — up or down.

The fault, however, isn’t in the keyword research itself. Keyword data is still the best tool we have for initial natural search optimization. The data collected from search queries is a window into human wants and needs.

But once optimization has begun and the first content modifications have gone live from keyword research, use stronger metrics based on actual performance.

Jill Kocher Brown
Jill Kocher Brown
Bio   •   RSS Feed